Immune checkpoint inhibitors (ICIs) have revolutionized cancer treatment. However, only some patients respond to ICIs, and current biomarkers for ICI efficacy have limited performance. Here, we devised an interpretable machine learning (ML) model trained using patient-specific cell-cell communication networks (CCNs) decoded from the patient’s bulk tumor transcriptome. The model could (i) predict ICI efficacy for patients across four cancer types (median AUROC: 0.79) and (ii) identify key communication pathways with crucial players responsible for patient response or resistance to ICIs by analyzing more than 700 ICI-treated patient samples from 11 cohorts. The model prioritized chemotaxis communication of immune-related cells and growth factor communication of structural cells as the key biological processes underlying response and resistance to ICIs, respectively. We confirmed the key communication pathways and players at the single-cell level in patients with melanoma. Our network-based ML approach can be used to expand ICIs’ clinical benefits in cancer patients.
The composition and diversity of gut microbiota significantly influence the immune system and are linked to various diseases, including inflammatory and allergy disorders. While considerable research has focused on exploring single bacterial species or consortia, the optimal strategies for microbiota-based therapeutics remain underexplored. Specifically, the comparative effectiveness of bacterial consortia versus individual species warrants further investigation. In our study, we assessed the impact of the bacterial consortium MPRO, comprising Lactiplantibacillus plantarum HY7712, Bifidobacterium animalis ssp. lactis HY8002, and Lacticaseibacillus casei HY2782, in comparison to its individual components. The administration of MPRO demonstrated enhanced therapeutic efficacy in experimental models of atopic dermatitis and inflammatory colitis when compared to single strains. MPRO exhibited the ability to dampen inflammatory responses and alter the gut microbial landscape significantly. Notably, MPRO administration led to an increase in intestinal CD103+CD11b+ dendritic cells, promoting the induction of regulatory T cells and the robust suppression of inflammation in experimental disease settings. Our findings advocate the preference for bacterial consortia over single strains in the treatment of inflammatory disorders, carrying potential clinical relevance.
Mouse models have been engineered to reveal the biological mechanisms of human diseases based on an assumption. The assumption is that orthologous genes underlie conserved phenotypes across species. However, genetically modified mouse orthologs of human genes do not often recapitulate human disease phenotypes which might be due to the molecular evolution of phenotypic differences across species from the time of the last common ancestor. Here, we systematically investigated the evolutionary divergence of regulatory relationships between transcription factors (TFs) and target genes in functional modules, and found that the rewiring of gene regulatory networks (GRNs) contributes to the phenotypic discrepancies that occur between humans and mice. We confirmed that the rewired regulatory networks of orthologous genes contain a higher proportion of species-specific regulatory elements. Additionally, we verified that the divergence of target gene expression levels, which was triggered by network rewiring, could lead to phenotypic differences. Taken together, a careful consideration of evolutionary divergence in regulatory networks could be a novel strategy to understand the failure or success of mouse models to mimic human diseases. To help interpret mouse phenotypes in human disease studies, we provide quantitative comparisons of gene expression profiles on our website (http://sbi.postech.ac.kr/w/RN).
26 September 2025 in Genomics, Proteomics & Bioinformatics
Diverse molecular networks have been extensively studied to discover therapeutic targets and repurpose approved drugs. However, it is necessary to select a suitable network since the performance of network medicine relies heavily on the completeness and characteristics of the selected network. Although a network using gene essentiality in cancer cells could be an effective platform for identifying anticancer targets, efforts to apply these networks to therapeutic applications have been limited. We constructed a phenotype-level network using co-essentiality relationships among genes from CRISPR screens across 769 cancer cell lines to discover therapeutic targets for diverse cancer types. By leveraging cancer driver genes and network propagation, we found that the co-essentiality network better prioritized anticancer targets and biomarkers and predicted more precise drug responses in cancer cells than other molecular networks. The co-essentiality network outperformed conventional molecular networks in drug repurposing and was validated in silico by clinical trial records. Notably, the co-essentiality network identified 30 repurposed drugs that the other networks have not yet covered, and we showcased three approved drugs repurposed for lung adenocarcinoma (atovaquone, eflornithine, and teriflunomide). Our study provides a novel network for precision oncology to improve the identification of therapeutic targets in specific cancers.
Diverse molecular networks have been extensively studied to discover therapeutic targets and repurpose approved drugs. However, it is necessary to select a suitable network since the performance of network medicine relies heavily on the completeness and characteristics of the selected network. Although a network using gene essentiality from cancer cells could be an effective platform for identifying anticancer targets, efforts to apply these networks in therapeutic applications have been limited. We constructed a phenotype-level network using the co-essentiality relationship between genes in CRISPR screens across 769 cancer cells to discover therapeutic targets for diverse cancer types. Leveraging cancer driver genes and network propagation on the networks, we found that the co-essentiality network better prioritized anticancer targets and biomarkers and predicted more precise drug responses in cancer cells than other molecular networks. The co-essentiality network outperformed conventional molecular networks in drug repurposing, which were validated in silico by clinical trial records. Notably, the co-essentiality network provided 37 repurposed drugs that the other networks have yet to cover, and we showcased three approved drugs repurposed for lung adenocarcinoma (Pioglitazone hydrochloride, Atovaquone, and Eflornithine). Our study provides a novel network for precision oncology to improve the identification of therapeutic targets in specific cancers.
Systemic lupus erythematosus (SLE) is a complex autoimmune disease, making accurate diagnosis and effective treatment challenging. Despite the critical nee